Endothelial cells (ECs) display marked phenotypic heterogeneity in structure and function, in time and space and in health and disease. Phenotypes differ between: 1) different segments of the vasculature (i.e., artery vs. vein vs. capillary); 2) different organs (i.e., lung capillary vs. heart capillary); and 3) neighboring cell (so-called mosaic heterogeneity). The mechanisms underlying the first two types of heterogeneity involve a combination of site-specific environmental signals (nurture) and vascular bed-specific epigenetic programs (nature). However, such mechanisms do not adequately explain mosaic heterogeneity of the endothelium, where phenotypically distinct neighbors have near-identical histories and environments. Mosaics of mutually exclusive phenotypes are observed across all domains of life. They are believed to arise from multistable regulatory circuits that are sensitive to biological noise. We propose that the normal endothelium leverages intrinsic noise-driven mosaic heterogeneity as a means of responding to opposing environmental demands with flexibility and speed. This ability to enhance its response repertoire (and hedge its bets) may be compromised in diseased or dysfunctional endothelium. The objective of this proposal is to use a combination of experimentation and modeling to delineate the mechanisms that govern mosaic heterogeneity of the endothelial-restricted procoagulant, von Willebrand factor (VWF), and to understand the adaptive significance of this phenomenon. We have shown that VWF expression is expressed in a minority of capillary ECs in heart, skeletal muscle, lung and brain and that this mosaic pattern is dynamically regulated. We have also observed dynamic mosaic heterogeneity of VWF in cultured primary human ECs. Surprisingly, we found that the bistable switch responsible for mosaic VWF expression involves noise-sensitive DNA methylation of the VWF promoter. Finally, and most importantly, our preliminary data indicate that the absence of mosaic VWF expression in the heart leads to profound endothelial damage. Based on these findings, we hypothesize that organ-specific mosaic heterogeneity of VWF is mediated by a cell-autonomous, bistable DNA methylation switch, whose sensitivity to noise is modulated by its extracellular and intracellular environment. We further hypothesize that mosaic VWF heterogeneity is required for cardiac health. Our goal is to measure, model and predict the behavior of this switch and to delineate the mechanisms by which mosaic heterogeneity promotes cardiac health. We propose three specific aims. In the first aim, we will delineate the role of biological noise in generating VWF mosaic heterogeneity. In the second aim, we will delineate the molecular basis of the bistable switch underlying VWF mosaic heterogeneity. In the third aim, we will delineate the role of VWF mosaic heterogeneity in cardiac health. The project is significant because it establishes a foundation for testing the hypothesis that dynamic mosaic heterogeneity of the endothelium provides a means for tunable, organ-specific bet hedging. Such behavior has important therapeutic implications for treating vascular diseases.